Method of detecting engine catastrophic risk during fracturing

Abstract

Some implementations include a method comprising training one or more risk models within a digital twin simulation system, wherein the digital twin simulation system is configured to model at least a first property of one or more hydraulic fracturing engines. The method may further comprise operating the one or more hydraulic fracturing engines, determining, at a first moment during operation, whether the one or more hydraulic fracturing engines are at risk of a ventilation failure, determining one or more actions to prevent the ventilation failure, and executing, via an edge control system, the one or more actions.

Description

TECHNICAL FIELD

[0001] The disclosure generally relates to the field of hydraulic fracturing, and more specifically, to anomaly detection in engines used for hydraulic fracturing operations.BACKGROUND

[0002] When fracturing with diesel pumps, the diesel engines may have a spun bearing or a seizure around the crankshaft journal during operation which may lead to catastrophic ventilation, severe damage to engines and vehicles, safety events involving personnel, etc. Therefore, the ability to detect elevated risk events before engine ventilation may be a significant advancement for business and Health, Safety, and Environment (HSE) initiatives.

[0003] It may be common practice to predict catastrophic engine ventilation using certain static measurements that deviate from normal behaviors. However, there may be many false positive predictions with static measurements which may interrupt fracturing operations and lead to downtime. Therefore, real-time risk monitoring techniques that use dynamic measurements may improve catastrophic event detection in diesel engines used for fracturing.BRIEF DESCRIPTION OF THE DRAWINGS

[0004] Implementations of the disclosure may be better understood by referencing the accompanying drawings.

[0005] FIG. 1 is an illustration depicting an example fracturing spread including an anomaly detection system, according to some implementations.

[0006] FIG. 2 is an illustration depicting an example flow diagram to be performed by the anomaly detection system, according to some implementations.

[0007] FIG. 3 is an illustration depicting an example computer, according to some implementations.

[0008] FIG. 4 is a flowchart depicting an example method of operations, according to some implementations.

[0009] FIGS. 1-4 and the operations described herein are examples meant to aid in understanding example implementations and should not be used to limit the potential imple0.mentations or limit the scope of the claims. None of the implementations described herein may be performed exclusively in the human mind nor exclusively using pencil and paper. None of the implementations described herein may be performed without computerized components such as those described herein. Some implementations may perform additional operations, fewer operations, operations in parallel or in a different order, and some operations differently.

[0010] The description that follows includes example systems, methods, techniques, and program flows that embody implementations of the disclosure. However, it is understood that this disclosure may be practiced without these specific details. In other instances, well-known instruction instances, protocols, structures, and techniques have not been shown in detail in order not to obfuscate the description.

[0011] Implementations of the machine-readable media, systems, and / or methods as described herein provide one or more improvements in the existing technology in the fields of anomaly detection and real-time monitoring and intervention. For example, the various implementations as described herein provide improvement(s) to existing technological processes by optimizing the management of engine ventilation risk of hydraulic fracturing engines. Traditional monitoring approaches, such as single-variable monitoring systems, may flag excessive instances of engine ventilation risk, leading to increased downtime. Other traditional monitoring approaches may also fail to identify multi-variate signifiers of engine ventilation risk, leading to potential engine ventilation failures. Various implementations as described herein also provide an improvement to the functioning of a computer by providing an algorithm in which, when implemented by the computer, may allow the computer to determine whether a hydraulic fracturing engine is at risk of ventilation. The algorithm may be updated using parallel processing pipelines, and the algorithm may minimize the number of false positives of engine ventilation risk detection through a multivariate, multi-scaled approach.Overview

[0012] To prevent catastrophic engine failures, improved techniques may be introduced to continuously monitor the risk of engine ventilation based on digital twin modeling. If an imminent or future risk of engine ventilation failure is detected, actions, which may include shutdown, may be taken to prevent engine ventilation and the loss of the engine. The engine risk models may be built on a digital twin platform with historical data for different types of engines. In some implementations, a ventilation failure may comprise an explosion of at least a portion of the hydraulic fracturing engine, a seizure of the fracturing engine, etc. This may result in a blown engine block, rendering the diesel engine inoperable and presenting a safety risk to onsite personnel.

[0013] In the improved technique, dynamic measurements may be used rather than static measurements to calculate the trend of variables at multi-timescales using real-time risk monitoring. The dynamic measurements may also describe how fast the engines deviate from normal behaviors. The improved technique may be based on a risk model that monitors multiple sensor data, such as the engine horsepower, engine revolutions per minute (RPM), oil pressure, engine load, etc. The engine risk model may be trained by the historical engine data on the digital twin platform. During training, the engine risk model may maximize the detection of true positive engine ventilation failures while minimizing the occurrence rate of false positive detections. The engine risk model, trained on the digital twin platform, may be mirrored to an edge system deployed in the field and may be run to monitor and protect pumping engines in real time. On average, engine ventilation may occur once per month for frac crews. The improved technique running on edge may potentially reduce the number of engine ventilation occurrences by 90%, thus delivering safer fracturing jobs to operators and saving maintenance and repair costs due to violent engine ventilation.Example Implementations

[0014] A digital twin simulation system and an edge control system may be used to monitor and control (if necessary) at least one diesel engine of a pump used in a hydraulic fracturing spread. A technique for anomaly detection in the diesel engine may utilize two parallel pipelines which work together to compute results. The digital twin simulation platform may train risk models with historical data. Meanwhile, the edge control system may run the trained model on edge in real time. The technique, described with additional detail below, may involve the following steps, respectively on the edge control system and digital twin platform:

[0015] 1. Determine the type of diesel engines on digital twin platform and edge control system

[0016] 2. Collect several periods of data prior to the target moment on digital twin platform and edge control system

[0017] 3. Determine if the fracturing engine unit is at risk of ventilation on digital twin platform and edge control system

[0018] 4. If yes, determine action(s) to prevent engine failure and execute the actions(s) on edge control system.

[0019] In some implementations, the type of diesel engine, as determined in step one, may be determined by reading data both on the digital twin platform and edge control system. In some implementations, several windows of time series data, according to step two, may be collected for analysis prior to a certain moment on both the digital twin platform and edge control system. The length of the time series window may range from 16 seconds to 1,024 seconds with a doubled length interval for critical variables such as oil pressure, engine load, engine revolutions per minute (RPM), etc.

[0020] In step three, the risk model may be represented by a risk score. The risk model may include three categories of scoring:

[0021] a) Oil leak: oil pressure may drop in the event of an oil leak in all windows.

[0022] b) Engine seizure and other mechanical problems: oil pressure may spike during oil block in short time windows.

[0023] c) Load surge: the engine load may surge when having a spun bearing or rod failure in short windows.

[0024] Accordingly, each window may only score for any of the above categories at any moment when satisfying a certain rule. The score of each category may equal the sum of scores from all time windows and may have an upper limit. The final risk score across all categories may equal the sum of scores from all categories. If the final score is above a certain threshold for a certain quantity of time, this may trigger alarms that the hydraulic fracturing engine(s) is / are at risk of ventilation failure. Accordingly, actions to remediate the failure may be taken in step four.

[0025] The anomaly detection system may identify whether an engine is at risk of ventilation using a multivariate, multi-layered approach. By design, the anomaly detection system may reach the threshold only when at least two categories (i.e., at least two variables) are triggered. For example, if an oil leak score at a certain moment is considered abnormal, but other categories are normal, then the anomaly detection system may not trigger any alarms. This may avoid unnecessary shutdowns, as multiple variables are considered when monitoring for ventilation risk. The multivariate, multi-layered approach may refer to using multiple different variables in anomaly detection. The multivariate, multi-layered approach may monitor both the physical action that has happened to / at the equipment (i.e., the fracturing engines) and the results of the physical action based on measured data. At least a portion of the measured properties may be calculated variables determined via an input function (such as engine load which may be calculated from an air / fuel calculation) and instrumentation variables (such as oil pressure which may be measured by a pressure gauge).

[0026] In one example, one or more properties of the hydraulic fracturing engine, such as oil pressure behavior, physics of oil flow through the engine, pressures at various orifices within the engine, etc. may be monitored by the anomaly detection system. Temporal data may also be considered, such as when observing a rate of change of oil pressure over a time window. The effect of these properties on various components of the hydraulic fracturing engine may also be observed through measuring engine load, temperature, engine horsepower, and RPMs over time. Traditional approaches, which may rely on single-layer, single variable monitoring or single variable shutdown thresholds (without accounting for temporal data) may may lead to excessive false positives.

[0027] The anomaly detection system may also utilize a multi-scale approach in addition to the multi-layer (multivariate) approach. The multi-scale approach may refer to how different changes in engine behavior over time are considered when initiating a shutdown sequence of the engine. If an event of a single risk category from the above risk categories occurs (e.g., an engine load surge), a risk score may be scaled according to a first scale. However, if at least two of the categories are occurring and generate a risk score over a threshold amount of time, this may be scored on a different scale, and the anomaly detection system may use the final risk score generated from the at least two categories to determine whether the engine is at risk of ventilation. The multi-scale approach may limit the detection rate of false positive events of engine ventilation.

[0028] All parameters and thresholds may be trained on the digital twin platform to generate appropriate actions to mitigate or prevent engine ventilation failure. In some implementations, the parameters and thresholds may be trained in response to all known engine ventilation events in one year of historical data. While retaining all true positives and historical data, the risk model may be configured to reduce the number of false positives by accommodating some special occasions and introducing additional sensor data.

[0029] In step three, the edge control system may run the model trained on the digital twin platform and generate alarms of real time data when the real time data matches the risk model and scores above the final risk score threshold.

[0030] In step four, actions to mitigate engine ventilation risk via the edge control system may include:

[0031] a) Sending a warning message to users. Warning messages may include multiple levels depending on severity (e.g., thresholds)

[0032] b) Suggesting users' rates setpoint(s) of individual pumps which may potentially mitigate or prevent engine ventilation.

[0033] c) Instead of suggestion in b), automatically sending rate setpoint(s) to a supervisory control system and or intervening directly by shutting down individual pumping units.

[0034] The above processes outlined in steps one through four may be run continuously (e.g., every control tick) at scheduled time instances, may be triggered by other modules, activated by a human operator, etc. If a fracturing unit is at risk of engine ventilation, the process may be repeated until the risk is fully mitigated (e.g., the final risk score is below the threshold(s)).Example Illustrations

[0035] FIG. 1 is an illustration depicting an example fracturing spread including an anomaly detection system, according to some implementations. A system 100 may include a wellhead 102 that is connected to a wellbore. The wellbore (not shown) may be fluidically connected to one or more subsurface formations for the purpose of hydrocarbon recovery. Although FIG. 1 shows only one wellhead 102, there may be any suitable number of wellheads 102 and wells.

[0036] The wellhead 102 may be connected to a manifold 104 via piping 106. The piping 106 may include one or more pipes between the wellhead 102 and the manifold 104. Any of the components at the wellsite may include or otherwise be coupled with one or more sensors 103. The manifold 104 may include a plurality of valves 108 and various internal piping (not shown) for performing hydraulic fracturing operations. Any of the valves and components described herein may include or otherwise be coupled with one or more sensors of any suitable type.

[0037] The manifold 104 may be connected to one or more frac pumps 112. The frac pumps 112 may include sensors such as temperature sensors, pressure sensors, viscosity sensors, amperage sensors, voltage sensors, flow sensor, and any other suitable sensor type. Each respective frac pump 112 also may include a lubrication controller (show in FIG. 1 as “LC”) configured to control lubrication of the respective frac pump 112. Operations of the lubrication controller and other frac pump components are described herein in further detail (for example see description of FIG. 2).

[0038] The frac pumps 112 may inject fracturing fluid into the wellbore under specified pressures and at predetermined flow rates. Each pump may be indicative of a single, discrete pumping device, but could alternatively comprise multiple pumps included as part of a pump truck or other pumping platform. All of the frac pumps 112 may or may not be the same type, size, configuration, or from the same manufacturer. Rather, some or all of the frac pumps 112 may be unique in size, output capability, etc. Each frac pump of the frac pumps 112 may include or may be powered by a hydraulic fracturing engine. In some implementations, the hydraulic fracturing engines may comprise one or more diesel engines. However, other engine configurations may also be used.

[0039] The manifold 104 also may be connected to a blender 116 via piping 118. The blender 116 may be connected via piping 128 to one or more chemical containers 120, water containers 122, and acid containers 124. The blender 116 also may be connected to a sand conveyor 130, where the sand conveyor 130 may be connected to the container of fracturing sanders 132.

[0040] The system 100 also may contain a supervisory control system 134 configured to control one or more of the components of the system 100 including the frac pumps 112 and the engines used to power the frac pumps 112. In some implementations, the supervisory control system 134 may directly control the equipment in operations for hydraulic fracturing. However, the supervisory control system 134 may interact with various equipment controllers (not shown) and sensors to perform operations related to hydraulic fracturing. The supervisory control system 134 may be coupled with an anomaly detection system 140. The anomaly detection system 140 may include a digital twin simulation system used to simulate each engine used to power the frac pumps 112. The anomaly detection system 140 may also include an edge control system. The edge control system may output commands to the supervisory control system 134 to perform one or more operations. For example, the anomaly detection system 140 may output a command to the supervisory control system 134 to shut down a diesel engine of at least one of the frac pumps 112.

[0041] A flow diagram describing the above anomaly detection and remediation procedure is now described. FIG. 2 is an illustration depicting an example flow diagram 200 to be performed by the anomaly detection system 140, according to some implementations. The operations of the flow diagram 200 may be performed by any combination of hardware / software, etc. The flow diagram includes a digital twin simulation system 202 and an edge control system 204 which may be implemented by any combination of hardware / software, etc. and used to perform the operations described herein. Operations begin at block 206.

[0042] At block 206, the digital twin simulation system 202 may determine a type of diesel engine used for a fracturing operation. For example, a hydraulic fracturing engine may include be comprised of a diesel engine used to drive one or more pumps. Different diesel engines may use different fuel mixtures and comprise different parts and differing operating parameters. Thus, a property considered to be abnormal for a first type of diesel engine may be considered normal for a different type of engine. Different types of diesel engines may exhibit different patterns if they are at risk of ventilation, may comprise different working flows, a different sensor system, etc. Flow progresses to block 208.

[0043] At block 208, one or more risk models may be trained on the digital twin simulation system 202. For example, the one or more risk models may be trained on the digital twin simulation system 202 to generate appropriate actions to avert all known engine ventilation events in at least one year of historical data. The historical data may include at least one year of data for various types of diesel engines in various conditions. All instances of true positive engine ventilation failures may be retained in the historical data used to train the risk models of the digital twin simulation system 202. During training, the one or more risk models may be rewarded for accurate predictions of true positive engine ventilation failures and penalized for false positives. In some implementations, the digital twin simulation system 202 may be a cloud-based processing system. The one or more risk models may be trained on the digital twin simulation system 202 and output to the edge control system 204 after training has concluded.

[0044] While keeping all true positives and historical data, the risk model may be trained to reduce the number of false positives by accommodating some special occasions in the historical data and introducing additional sensor data. For example, one or more risk models may consider a special occasion such as gas substitution in the hydraulic fracturing engine. Some hydraulic fracturing engines may be dual-fuel systems configured to operate on a percentage of diesel fuel and a percentage of natural gas. Due to gas substitution, variables such as engine load and oil pressure may strongly fluctuate as the fuel mixture is changed despite the engine operating normally. Other contaminants may also be introduced into the engine which may alter sensor data when compared to engine operation using primarily diesel fuel. Considering special occasions such as gas substitution may eliminate a number of false positive identifications of ventilation failure. Other special occasions may also be possible.

[0045] The digital twin simulation system 202 and edge control system 204 may be operated as a parallel pipeline system. For example, a first pipeline may include simulated operations performed via the digital twin simulation system 202, such as training and updating the or more risk models of block 208 via cloud computing. A second pipeline may include data obtained and operations performed by the edge control system 204. The edge control system 204 may refer to a control system physically located closer to a user or measured device. For example, the edge control system 204 may be physically located at the well site having the frac pumps 112 for reduced latency during computations and outputs.

[0046] The two parallel pipelines may communicate with one another. For example, the updated model(s) of block 208 may be mirrored on edge devices, internet of things (IoT) systems, etc. within the edge control system 204 at block 214 to implement any changes. Updates may, for example, be based upon new historical data used to train the risk models of block 208. The new historical data may allow the digital twin simulation system 202 to model updated diesel engines. The one or more risk models may be retrained to predict ventilation failure based on one or more dynamic properties of the updated diesel engine. The updated model may then be pushed from the digital twin simulation system 202 to the edge control system 204 to identify engine ventilation risks and / or events that prior versions of the risk models may have missed. Flow may progress to block 214, where the trained risk models are output from the digital twin simulation system 202 to the edge control system 204. The edge control system 204 is now described. Operations of the edge control system 204 begin at block 210.

[0047] At block 210, the edge control system 204 may determine a type of diesel engine used during the fracturing operation. In some implementations, the type of diesel engine may be determined by reading data both on the digital twin simulation system 202 and the edge control system 204. Flow progresses to block 212.

[0048] At block 212, the edge control system 204 may collect several windows of data prior to a first current moment. Several windows of time series data prior to a certain moment may be collected for analysis via the digital twin simulation system 202 and the edge control system 204. The length of each time series window may range from 16 seconds to 1024 seconds with a doubled length interval for certain variables such as oil pressure, engine load, engine RPM, etc. Flow progresses to block 214.

[0049] At block 214, the edge control system 204 may determine whether the hydraulic fracturing engine is at risk of ventilation. The edge control system 204 may run the risk model trained on the digital twin platform and generate alarms of real time data that matches the risk model and scores above one or more thresholds. If the engine is at risk of ventilation, flow progresses to block 216. If not, flow of the flow diagram 200 may cease at block 220. In some implementations, the flow of the flow diagram 200 may return from block 220 to block 212, where the edge control system 204 may collect several windows of data prior to a second current moment which may be later in time than the first current moment of block 212.

[0050] The risk model(s) output by the digital twin simulation system 202 may quantify the severity of any detected risks to the health of the fracturing engine. In some implementations, the risk model may be represented by a risk score. The risk model may include three categories of scoring: oil leak, engine seizure and other mechanical problems, and load surge. Each category may be scored depending on the likelihood of the respective events' occurrences. For example, an oil leak may be detected (and scored) when oil pressure drops in all windows. Engine seizure and other mechanical problems may be detected and scored when oil pressure jumps during oil block in short time windows. A short time window may be defined as a time window between sixteen seconds and five minutes. A load surge may be detected and scored in the event measured engine load spiking. The engine load may surge when having a spun bearing or rod failure in short time windows.

[0051] At any moment, each window of block 212's windows of time series data may only score for any of the above categories if satisfying a certain rule. This rule may be a threshold value for each category sustained over the time window (e.g., an oil pressure below ten psi over a sixteen second time window). The score of every category may equal the sum of scores from all windows, and each category's score may have an upper limit. The final risk score may be the sum of scores from all categories. In some implementations, a rate of change of a variable over time may also affect a category's risk score; an oil pressure dropping from thirty psi to ten psi over a sixteen second period may score higher than an oil pressure reduction from thirty psi to ten psi over a four-minute period. Assuming the fracturing engine is determined by the risk models and edge control system 204 to be at risk of ventilation, flow progresses to block 216.

[0052] At block 216, the edge control system 204 may determine one or more actions to prevent engine ventilation of the one or more hydraulic fracturing engines. The edge control system may determine the one or more actions if the final score from block 214 is determined to be above a certain threshold for a certain amount of time. This may trigger alarms and the edge control system 204 for remedial action. By design, the edge control system 204 may reach the threshold only when at least two categories are triggered. Flow progresses to block 218.

[0053] At block 218, the edge control system 204 may execute the one or more actions determined at block 216. Actions to mitigate high engine risk using the edge control system 204 may include sending warning messages to users, suggesting users' rates setpoint(s) of individual pumps which may potentially mitigate or prevent engine ventilation, automatically sending rate setpoint(s) to a supervisory control system, etc. Should the edge control system 204 send warning messages to users, the warning messages may have multiple levels depending on severity (e.g., thresholds). In severe cases, the edge control system 204 may output commands to a supervisory control system, such as the supervisory control system 134, to directly shut down individual pumping units (such as the frac pumps 112) based on their final risk score. From block 218, flow progresses to block 220, where operations of the flow diagram 200 cease. If an engine unit is determined to be at continued risk of engine ventilation, at least a portion of processes of the flow diagram 200 may be repeated until the risk is fully mitigated (e.g., the risk score is below the threshold). From block 220, flow may return to block 212, where new windows of time series data may be collected via the edge control system 204.

[0054] As shown, the above techniques may utilize two parallel pipelines to work together. The digital twin simulation system 202 may train the one or more risk models with historical data. Meanwhile, the edge control system 204 may operate the trained one or more risk models in real time. At least a portion of the operations described in the flow diagram 200 may be performed by both the digital twin simulation system 202 and the edge control system 204. For example, determining the type of diesel engine may be performed both via the digital twin simulation system 202 and edge control system 204. Collecting several periods of data prior to a target moment may be performed both by the digital twin simulation system 202 and the edge control system 204. Determining whether the fracturing engine unit is at risk of ventilation may also be performed in parallel on both the digital twin simulation system 202 and the edge control system 204. If the fracturing engine is determined to be at risk of engine ventilation, one or more actions may be executed via the edge control system 204 to prevent engine failure.Example Computer

[0055] FIG. 3 is an illustration depicting an example computer 300, according to some implementations. The computer 300 may include a processor 301 (possibly including multiple processors, multiple cores, multiple nodes, and / or implementing multi-threading, etc.). The computer system may include memory 307. The memory 307 may be system memory or any one or more of the above already described possible realizations of machine-readable media. The computer system may also include a bus 303 and a network interface 305. The system may communicate via transmissions to and / or from remote devices via the network interface 305 in accordance with a network protocol corresponding to the type of network interface, whether wired or wireless and depending upon the carrying medium. In addition, a communication or transmission may involve other layers of a communication protocol and or communication protocol suites (e.g., transmission control protocol, Internet Protocol, user datagram protocol, virtual private network protocols, etc.).

[0056] The computer 300 may further include an anomaly detector 312, and the anomaly detector 312 may include an edge controller 314 and digital twin simulator 316. In some implementations, the anomaly detector 312 may be similar to FIG. 1's anomaly detection system 140, the edge controller 314 may be similar to the edge control system 204, and the digital twin simulator 316 may be similar to the digital twin simulation system 202. The anomaly detector 312, edge controller 314, and digital twin simulator 316 may be implemented in hardware, software, and / or other logic configured to perform the operations described herein. In some implementations, the anomaly detector 312 may be implemented as instructions executable on the processor 301. The anomaly detector 312 may include computerized functionality configured to generate an engine ventilation risk score using one or more risk models, output suggestions to a user interface, and output commands to one or more items of equipment to mitigate the ventilation risk. In some implementations, the anomaly detector 312 may output commands to the supervisory controller 310, and the supervisory controller 310 may directly control one or more items of equipment of the frac fleet, initiate a shutdown of one or more frac pumps, etc., to mitigate the ventilation risk. In some implementations, the edge controller 314 may include the supervisory controller 310. Therefore, the edge controller 314 may directly intervene and shut down one or more frac pumps directly to mitigate the risk of engine ventilation in severe scenarios. Other remedial measures may also be possible.

[0057] Any one of the previously described functionalities may be partially (or entirely) implemented in hardware and / or on the processor 301. For example, the functionality may be implemented with an application specific integrated circuit, in logic implemented in the processor 301, in a co-processor on a peripheral device or card, etc. Further, realizations may include fewer or additional components not illustrated in FIG. 3 (e.g., video cards, audio cards, additional network interfaces, peripheral devices, etc.). The processor 301 and the network interface 305 are coupled to the bus 303. Although illustrated as being coupled to the bus 303, the memory 307 may be coupled to the processor 301.Example Method of Operations

[0058] FIG. 4 is a flowchart depicting an example method of operations, according to some implementations. Operations of a method 400 may be performed by software, firmware, hardware, or a combination thereof. Such operations are described with reference to FIGS. 1-3. However, such operations may be performed by other systems or components. The operations of the method 400 begin at block 402.

[0059] At block 402, the method 400 includes training one or more risk models within a digital twin simulation system, wherein the digital twin simulation system is configured to model at least a first property of one or more hydraulic fracturing engines. Flow progresses to block 404.

[0060] At block 404, the method 400 includes operating the one or more hydraulic fracturing engines. Flow progresses to block 406.

[0061] At block 406, the method 400 includes determining, at a first moment during operation, whether the one or more hydraulic fracturing engines are at risk of a ventilation failure. Flow progresses to block 408.

[0062] At block 408, the method 400 includes determining one or more actions to prevent the ventilation failure. Flow progresses to block 410.

[0063] At block 410, the method 400 includes executing, via an edge control system, the one or more actions. For example, the edge control system 204 may send a warning message to a user, suggest rate setpoints of individual pumps, automatically output rate setpoints to a supervisory control system, directly shut down individual pumps, etc. Flow of the method 400 ceases.

[0064] While the aspects of the disclosure are described with reference to various implementations and exploitations, it will be understood that these aspects are illustrative and that the scope of the claims is not limited to them. In general, techniques for anomaly detection using an edge control system and digital twin platform operated in parallel as described herein may be implemented with facilities consistent with any hardware system or hardware systems. Many variations, modifications, additions, and improvements may be possible.

[0065] Plural instances may be provided for components, operations or structures described herein as a single instance. Finally, boundaries between various components, operations and data stores are somewhat arbitrary, and particular operations are illustrated in the context of specific illustrative configurations. Other allocations of functionality are envisioned and may fall within the scope of the disclosure. In general, structures and functionality presented as separate components in the example configurations may be implemented as a combined structure or component. Similarly, structures and functionality presented as a single component may be implemented as separate components. These and other variations, modifications, additions, and improvements may fall within the scope of the disclosure.

[0066] Use of the phrase “at least one of” preceding a list with the conjunction “and” should not be treated as an exclusive list and should not be construed as a list of categories with one item from each category, unless specifically stated otherwise. A clause that recites “at least one of A, B, and C” may be infringed with only one of the listed items, multiple of the listed items, and one or more of the items in the list and another item not listed. As an example, “at least one of: a, b, or c” is intended to cover: a, b, c, a-b, a-c, b-c, and a-b-c.

[0067] The various illustrative logics, logical blocks, modules, circuits, and algorithm processes described in connection with the implementations disclosed herein may be implemented as electronic hardware, computer software, or combinations of both. The interchangeability of hardware and software has been described generally, in terms of functionality, and illustrated in the various illustrative components, blocks, modules, circuits and processes described throughout. Whether such functionality is implemented in hardware or software depends upon the particular application and design constraints imposed on the overall system.

[0068] The hardware and data processing apparatus used to implement the various illustrative logics, logical blocks, modules and circuits described in connection with the implementations disclosed herein may be implemented or performed with a general purpose single-or multi-chip processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general-purpose processor may be a microprocessor or any conventional processor, controller, microcontroller, or state machine. A processor also may be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration. In some implementations, particular processes and methods may be performed by circuitry that is specific to a given function.

[0069] In one or more implementations, the functions described may be implemented in hardware, digital electronic circuitry, computer software, firmware, including the structures disclosed in this specification and their structural equivalents thereof, or in any combination thereof. Implementations of the subject matter described in this specification also may be implemented as one or more computer programs, e.g., one or more modules of computer program instructions stored on a computer storage media for execution by, or to control the operation of, a computing device.

[0070] If implemented in software, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. The processes of a method or algorithm disclosed herein may be implemented in processor-executable instructions which may reside on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that may be enabled to transfer a computer program from one place to another. Storage media may be any available media that may be accessed by a computer. By way of example, and not limitation, such computer-readable media may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that may be used to store desired program code in the form of instructions or data structures and that may be accessed by a computer. Also, any connection may be properly termed a computer-readable medium. Disk and disc, as used herein, includes compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk, and Blu-Ray™ disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations also may be included within the scope of computer-readable media. Additionally, the operations of a method or algorithm may reside as one or any combination or set of codes and instructions on a machine readable medium and computer-readable medium, which may be incorporated into a computer program product.

[0071] Various modifications to the implementations described in this disclosure may be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other implementations without departing from the spirit or scope of this disclosure. Thus, the claims are not intended to be limited to the implementations shown herein but are to be accorded the widest scope consistent with this disclosure, the principles and the novel features disclosed herein.

[0072] Certain features that are described in this specification in the context of separate implementations also may be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation also may be implemented in multiple implementations separately or in any suitable sub-combination. Moreover, although features may be described as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination may in some cases be excised from the combination, and the claimed combination may be directed to a sub-combination or variation of a sub-combination.

[0073] While operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. Further, the drawings may schematically depict one more example process in the form of a flow diagram. However, some operations may be omitted and / or other operations that are not depicted may be incorporated in the example processes that are schematically illustrated. For example, one or more additional operations may be performed before, after, simultaneously, or between any of the illustrated operations. In certain circumstances, multitasking and parallel processing may be advantageous. Moreover, the separation of various system components in the implementations described should not be understood as requiring such separation in all implementations, and the described program components and systems may generally be integrated together in a single software product or packaged into multiple software products. Additionally, other implementations are within the scope of the following claims. In some cases, the actions recited in the claims may be performed in a different order and still achieve desirable results.

[0074] Unless otherwise specified, use of the terms “up,”“upper,”“upward,”“uphole,”“upstream,” or other like terms shall be construed as generally away from the bottom, terminal end of a well; likewise, use of the terms “down,”“lower,”“downward,”“downhole,” or other like terms shall be construed as generally toward the bottom, terminal end of the well, regardless of the wellbore orientation. Use of any one or more of the foregoing terms shall not be construed as denoting positions along a perfectly vertical axis. In some instances, a part near the end of the well may be horizontal or even slightly directed upwards. Unless otherwise specified, use of the terms “subsurface formation” or “subterranean formation” shall be construed as encompassing both areas below exposed earth and areas below earth covered by water such as ocean or fresh water.Example Implementations

[0075] Example implementations include the following:

[0076] Implementation #1: A method comprising: training one or more risk models within a digital twin simulation system, wherein the digital twin simulation system is configured to model at least a first property of one or more hydraulic fracturing engines; operating the one or more hydraulic fracturing engines; determining, at a first moment during operation, whether the one or more hydraulic fracturing engines are at risk of a ventilation failure; determining one or more actions to prevent the ventilation failure; and executing, via an edge control system, the one or more actions.

[0077] Implementation #2: The method of Implementation 1, further comprising: determining, via the edge control system, an engine type of the one or more hydraulic fracturing engines; and determining, within the digital twin simulation system, the engine type of the one or more hydraulic fracturing engines.

[0078] Implementation #3: The method of any one or more of Implementations 1-2, wherein the one or more hydraulic fracturing engines comprise one or more diesel engines, and wherein determining the engine type comprises determining a diesel engine type of the one or more hydraulic fracturing engines.

[0079] Implementation #4: The method of any one or more of Implementations 1-3, wherein the one or more risk models are output to the edge control system after the training is complete.

[0080] Implementation #5: The method of any one or more of Implementations 1-4, further comprising: collecting, via the edge control system, one or more windows of time series data prior to the first moment.

[0081] Implementation #6: The method of any one or more of Implementations 1-5, further comprising: generating, via the one or more risk models at the first moment, a risk score for each category of one or more risk categories, wherein the one or more risk categories include an oil leak category, load surge category, and engine seizure category; and generating a risk score of the one or more hydraulic fracturing engines when at least two of the one or more risk categories exceed a threshold score over a threshold period of time.

[0082] Implementation #7: The method of any one or more of Implementations 1-6, wherein executing, via the edge control system, the one or more actions comprises at least one of outputting a warning to a user, outputting rate setpoint suggestions to the user, outputting rate setpoint commands to a supervisory control system, or initiating a shutdown of the one or more hydraulic fracturing engines.

[0083] Implementation #8: An anomaly detection system comprising: a digital twin simulation system configured to model at least a first property of one or more hydraulic fracturing engines; an edge control system coupled with the one or more hydraulic fracturing engines; and a computer-readable medium having instructions executable by the one or more processors, the instructions including: instructions to train one or more risk models within the digital twin simulation system, instructions to operate the one or more hydraulic fracturing engines, instructions to determine, at a first moment during operation, whether the one or more hydraulic fracturing engines are at risk of a ventilation failure, instructions to determine one or more actions to prevent the ventilation failure, and instructions to execute, via the edge control system, the one or more actions.

[0084] Implementation #9: The anomaly detection system of Implementation 8, further comprising: instructions to determine, via the edge control system, an engine type of the one or more hydraulic fracturing engines; and instructions to determine, within the digital twin simulation system, the engine type of the one or more hydraulic fracturing engines.

[0085] Implementation #10: The anomaly detection system of any one or more of Implementations 8-9, wherein the one or more hydraulic fracturing engines comprise one or more diesel engines, and wherein the instructions to determine the engine type comprise instructions to determine a diesel engine type of the one or more hydraulic fracturing engines.

[0086] Implementation #11: The anomaly detection system of any one or more of Implementations 8-10, further comprising: instructions to deploy the one or more risk models to the edge control system after the training is complete.

[0087] Implementation #12: The anomaly detection system of any one or more of Implementations 8-11, further comprising: instructions to collect, via the edge control system, one or more windows of time series data of at least the first property prior to the first moment.

[0088] Implementation #13: The anomaly detection system of any one or more of Implementations 8-12, further comprising: instructions to generate, via the one or more risk models at the first moment, a risk score for each category of one or more risk categories, wherein the one or more risk categories include an oil leak category, load surge category, and engine seizure category; and instructions to generate a risk score of the one or more hydraulic fracturing engines when at least two of the one or more risk categories exceed a threshold score over a threshold period of time.

[0089] Implementation #14: The anomaly detection system of any one or more of Implementations 8-13, wherein the instructions to execute, via the edge control system, the one or more actions comprise instructions to output a warning to a user, instructions to output rate setpoint suggestions to the user, instructions to output rate setpoint commands to a supervisory control system, or instructions to initiate a shutdown of the one or more hydraulic fracturing engines.

[0090] Implementation #15: One or more non-transitory machine-readable media including instructions executable by one or more processors, the instructions comprising: instructions to train one or more risk models within a digital twin simulation system, wherein the digital twin simulation system is configured to model at least a first property of one or more hydraulic fracturing engines; instructions to operate the one or more hydraulic fracturing engines; instructions to determine, at a first moment during operation, whether the one or more hydraulic fracturing engines are at risk of a ventilation failure; instructions to determine one or more actions to prevent the ventilation failure; and instructions to execute, via an edge control system, the one or more actions.

[0091] Implementation #16: The machine-readable media of Implementation 15, further comprising: instructions to determine, via the edge control system, an engine type of the one or more hydraulic fracturing engines; and instructions to determine, within the digital twin simulation system, the engine type of the one or more hydraulic fracturing engines.

[0092] Implementation #17: The machine-readable media of any one or more of Implementations 15-16, wherein the one or more hydraulic fracturing engines comprise one or more diesel engines, and wherein the instructions to determine the engine type comprise instructions to determine a diesel engine type of the one or more hydraulic fracturing engines.

[0093] Implementation #18: The machine-readable media of any one or more of Implementations 15-17, further comprising: instructions to collect, via the edge control system, one or more windows of time series data of at least the first property prior to the first moment.

[0094] Implementation #19: The machine-readable media of any one or more of Implementations 15-18, further comprising: instructions to generate, via the one or more risk models at the first moment, a risk score for each category of one or more risk categories, wherein the one or more risk categories include an oil leak category, load surge category, and engine seizure category; and instructions to generate a risk score of the one or more hydraulic fracturing engines when at least two of the one or more risk categories exceed a threshold score over a threshold period of time.

[0095] Implementation #20: The machine-readable media of any one or more of Implementations 15-19, wherein the instructions to execute, via the edge control system, the one or more actions comprise instructions to output a warning to a user, instructions to output rate setpoint suggestions to the user, instructions to output rate setpoint commands to a supervisory control system, or instructions to initiate a shutdown of the one or more hydraulic fracturing engines.

Claims

1. A method comprising:training one or more risk models within a digital twin simulation system, wherein the digital twin simulation system is configured to model at least a first property of one or more hydraulic fracturing engines;operating the one or more hydraulic fracturing engines;determining, at a first moment during operation, whether the one or more hydraulic fracturing engines are at risk of a ventilation failure;determining one or more actions to prevent the ventilation failure; andexecuting, via an edge control system, the one or more actions.

2. The method of claim 1, further comprising:determining, via the edge control system, an engine type of the one or more hydraulic fracturing engines; anddetermining, within the digital twin simulation system, the engine type of the one or more hydraulic fracturing engines.

3. The method of claim 2, wherein the one or more hydraulic fracturing engines comprise one or more diesel engines, and wherein determining the engine type comprises determining a diesel engine type of the one or more hydraulic fracturing engines.

4. The method of claim 1, wherein the one or more risk models are output to the edge control system after the training is complete.

5. The method of claim 1, further comprising:collecting, via the edge control system, one or more windows of time series data prior to the first moment.

6. The method of claim 1, further comprising:generating, via the one or more risk models at the first moment, a risk score for each category of one or more risk categories, wherein the one or more risk categories include an oil leak category, load surge category, and engine seizure category; andgenerating a risk score of the one or more hydraulic fracturing engines when at least two of the one or more risk categories exceed a threshold score over a threshold period of time.

7. The method of claim 1, wherein executing, via the edge control system, the one or more actions comprises at least one of outputting a warning to a user, outputting rate setpoint suggestions to the user, outputting rate setpoint commands to a supervisory control system, or initiating a shutdown of the one or more hydraulic fracturing engines.

8. An anomaly detection system comprising:a digital twin simulation system configured to model at least a first property of one or more hydraulic fracturing engines;an edge control system coupled with the one or more hydraulic fracturing engines; anda computer-readable medium having instructions executable by one or more processors, the instructions including:instructions to train one or more risk models within the digital twin simulation system,instructions to operate the one or more hydraulic fracturing engines,instructions to determine, at a first moment during operation, whether the one or more hydraulic fracturing engines are at risk of a ventilation failure,instructions to determine one or more actions to prevent the ventilation failure, andinstructions to execute, via the edge control system, the one or more actions.

9. The anomaly detection system of claim 8, further comprising:instructions to determine, via the edge control system, an engine type of the one or more hydraulic fracturing engines; andinstructions to determine, within the digital twin simulation system, the engine type of the one or more hydraulic fracturing engines.

10. The anomaly detection system of claim 9, wherein the one or more hydraulic fracturing engines comprise one or more diesel engines, and wherein the instructions to determine the engine type comprise instructions to determine a diesel engine type of the one or more hydraulic fracturing engines.

11. The anomaly detection system of claim 8, further comprising:instructions to deploy the one or more risk models to the edge control system after the training is complete.

12. The anomaly detection system of claim 8, further comprising:instructions to collect, via the edge control system, one or more windows of time series data of at least the first property prior to the first moment.

13. The anomaly detection system of claim 8, further comprising:instructions to generate, via the one or more risk models at the first moment, a risk score for each category of one or more risk categories, wherein the one or more risk categories include an oil leak category, load surge category, and engine seizure category; andinstructions to generate a risk score of the one or more hydraulic fracturing engines when at least two of the one or more risk categories exceed a threshold score over a threshold period of time.

14. The anomaly detection system of claim 8, wherein the instructions to execute, via the edge control system, the one or more actions comprise instructions to output a warning to a user, instructions to output rate setpoint suggestions to the user, instructions to output rate setpoint commands to a supervisory control system, or instructions to initiate a shutdown of the one or more hydraulic fracturing engines.

15. One or more non-transitory machine-readable media including instructions executable by one or more processors, the instructions comprising:instructions to train one or more risk models within a digital twin simulation system, wherein the digital twin simulation system is configured to model at least a first property of one or more hydraulic fracturing engines;instructions to operate the one or more hydraulic fracturing engines;instructions to determine, at a first moment during operation, whether the one or more hydraulic fracturing engines are at risk of a ventilation failure;instructions to determine one or more actions to prevent the ventilation failure; andinstructions to execute, via an edge control system, the one or more actions.

16. The machine-readable media of claim 15, further comprising:instructions to determine, via the edge control system, an engine type of the one or more hydraulic fracturing engines; andinstructions to determine, within the digital twin simulation system, the engine type of the one or more hydraulic fracturing engines.

17. The machine-readable media of claim 16, wherein the one or more hydraulic fracturing engines comprise one or more diesel engines, and wherein the instructions to determine the engine type comprise instructions to determine a diesel engine type of the one or more hydraulic fracturing engines.

18. The machine-readable media of claim 15, further comprising:instructions to collect, via the edge control system, one or more windows of time series data of at least the first property prior to the first moment.

19. The machine-readable media of claim 15, further comprising:instructions to generate, via the one or more risk models at the first moment, a risk score for each category of one or more risk categories, wherein the one or more risk categories include an oil leak category, load surge category, and engine seizure category; andinstructions to generate a risk score of the one or more hydraulic fracturing engines when at least two of the one or more risk categories exceed a threshold score over a threshold period of time.

20. The machine-readable media of claim 15, wherein the instructions to execute, via the edge control system, the one or more actions comprise instructions to output a warning to a user, instructions to output rate setpoint suggestions to the user, instructions to output rate setpoint commands to a supervisory control system, or instructions to initiate a shutdown of the one or more hydraulic fracturing engines.

Images